Pattern recognition algorithms and disease biomarkers

Ben A. Hitt, Emanuel Petricoin, Lance Liotta

Research output: Chapter in Book/Report/Conference proceedingChapterpeer-review


The publication of our successes1-4 in using mass spectral patterns to identify patients with ovarian and prostate cancers and animals with premalignant pancreatic cancer has led to increased interest in using these methods to characterize a variety of disease and biological states. Adoption of a pattern discovery and recognition approach to biomarker identification requires the acceptance of two basic premisesthat there is a need and that a pattern of features in a signal constitute a biomarkeras well as choosing an adequate set of tools. This chapter will present a case for acceptance of both premises, describe some of the methods available, and present the pitfalls associated with the process.

Original languageEnglish
Title of host publicationInformatics in Proteomics
PublisherCRC Press
Number of pages14
ISBN (Electronic)9781420027624
ISBN (Print)9781574444803
StatePublished - 1 Jan 2005
Externally publishedYes


Dive into the research topics of 'Pattern recognition algorithms and disease biomarkers'. Together they form a unique fingerprint.

Cite this